Hi Tony,
IMHO - It's not easy. DNN development is a tricky beast - a small bug can take hours. I was unable to cope with the number of requests/support calls and demands from customers (customer management and expectation is critical).
My hot tip - IF you do decide to do this - pick a simple "utility" type module. Don't make the mistake that I did and pick "a social networking suite" to rival Facebooks ;) It's not possible for one person or a small team to develop such complex software, unless they have a few years to spare. It also has an infinite roadmap - not wise.
However, a small utility type module that has a small roadmap can be a moneyspinner. For example - I use DNNStuff's Aggregator mod all over the place. Richard only has to do a release a few times a year and there are few support calls (you could argue this is the quality of the product, and of course this is a factor), however simplicity is key.
It's hard to quantify how many sales you'll make. It depends on the niche and the compeition. For exampel, BizModules picked an untapped area (Videos/gallery) and ran with it. They have made a LOT of money - very clever. ActiveModules has the forum market.
Up until a little while ago, a good Google map module would have made a LOT of money - now Netism is cornering that side of things. Pick a niche area where there is a need and go for it. But keep it simple ;)
The other 2 areas you HAVE to consider - correct licensing/pricing (don't price yourself out the market - this was one of my key mistakes and the cause of the demise of Smart-Thinker as a commerical vendor) and manage your customer expectations.
Licensing/Pricing strategy would be a 10 page post. My advice would be to look at how one of the successfuly vendors do it (ActiveForums/Ifinity). Ventrian is another successful module publisher but WAY to cheap IMHO - should be at least $100 for a 6 complex mod subscription (note- these are purely my speculations to survive).
For some reasons customers expect all change requests and features tomorrow. It's not possible. By all means take suggestions from customers but publish a simple roadmap, release early and often and keep it all flexible. Communicate as often as possible (again, time...) and set the standard early as to how you go about your business.
And lastly: I have been involved in a few markets and there is a limited period of time for one-man developers (there are always exceptions of course). For example, when the first PocketPCs came out I used to sell a game I wrote in EmbeddedVB (and later converted to .NET) - but in a year companies like EA started their own mobile games publishing business and I could not compete against teams. The level of professionalism and evolution of dev teams (as opposed to single developers) is advancing fast - I believe the days of single developers are numbered...
Feel free to prove me wrong ;) - Good luck!